B cells mediate humoral immunity by producing specific antibodies. The basic structural subunit of an antibody (Ab) is an immunoglobulin (Ig) molecule. Ig molecules consist of a complex of two identical heavy (H) and two identical light (L) polypeptide chains. At the amino terminus of each H chain and L chain is a region that varies in amino acid sequence named the variable (V) region. The remaining portion of the H and L chains is relatively constant in amino acid sequence and is named the constant (C) region. In an Ig molecule, the H and L chain V regions (VH and VL) are juxtaposed to form the potential antigen-binding site. The genes that encode H and L chain V regions are assembled somatically from segments of germline DNA during precursor B (pre-B) cell differentiation: V, D and J gene segments for the H chain and V and J gene segments for the L chain. Within Ig V regions are three regions of greatest amino acid sequence variability that interact to form the antigen-recognition site and are thus referred to as complementarity determining regions (CDRs).
The V gene segment encodes the bulk of the V region domain, including CDR1 and CDR2. Diversity in CDR1 and CDR2 derives from sequence heterogeneity among multiple different germline-encoded V segments. CDR3 is encoded by sequences that are formed by the joining of H chain V, D, and J gene segments and L chain V and J segments and by mechanisms that create nucleotide sequence heterogeneity where these segments are combined. Additional diversity may be derived from pairing of different H and L chain V regions. Collectively these processes yield a primary repertoire of antibodies encoded by germline gene segments and expressed by newly formed B cells.
An additional source of antibody diversity is imposed on top of the diversity generated by recombination of Ig gene segments. B cells are able to introduce mutations into the antibody V regions that they express, a process called somatic hypermutation. Thus, when an animal first encounters an antigen, the antigen binds to a specific B cell which happens to carry antibodies which have a V domain which binds the antigen. This primary response may activate this B cell to go on to secrete the cognate antibody. These activated B cells can also now target a somatic mutation process to their rearranged antibody gene segments and thus allow the production of daughter cells which make variants of the antibodies of the primary response. A selection process amplifies those variant B cell descendants which make an antibody of improved affinity of the antigen. In B cells, somatic hypermutations are targeted to a restricted genomic region including both the rearranged VH and VL genes. Thus somatic mutation allows affinity maturation—the production and selection of high affinity antibodies. Therefore, somatic mutation is important for the generation of high affinity antibodies.
The exquisite specificity and high affinity of antibodies and the discovery of hybridoma technology allowing the generation of monoclonal antibodies (mAbs) has generated great expectations for their utilization as targeted therapeutics for human diseases. MAbs are identical because they are produced by a single B cell and its progeny. MAbs are made by fusing the spleen cells from a mouse that has been immunized with the desired antigen with myeloma cells to generate immortalized hybridomas. One of the major impediments facing the development of in vivo applications for mAbs in humans is the intrinsic immunogenicity of non-human Igs. Patients respond to therapeutic doses of mouse mAbs by making antibodies against the mouse Ig sequences (Human Anti Mouse Antibodies; HAMA), causing acute toxicity, alter their biodistribution and accelerate clearance, thus reducing the efficacy of subsequent administrations (Mirick et al. (2004), Q. Nucl. Med. Mol. Imaging 48:251-257).
To circumvent the generation of HAMA, antibody humanization methods have been developed in an attempt to produce mAbs with decreased immunogenicity when applied to humans. These endeavors have yielded various recombinant DNA-based approaches aimed at increasing the content of human amino acid sequences in mAbs while retaining the specificity and affinity of the parental non-human antibody. Humanization began with the construction of mouse-human chimeric mAbs (S. L. Morrison et al. (1984), Proc. Natl. Acad. Sci. USA 81:6851-5), in which the Ig C regions in murine mAbs were replaced by human C regions. Chimeric mAbs contain 60-70% of human amino acid sequences and are considerably less immunogenic than their murine counterparts when injected into humans, albeit that a human anti-chimeric antibody response was still observed (W. Y. Hwang et al. (2005), Methods 36:3-10).
In attempts to further humanize murine mAbs, CDR grafting was developed. In CDR grafting, murine antibodies are humanized by grafting their CDRs onto the VL and VH frameworks of human Ig molecules, while retaining those murine framework residues deemed essential for specificity and affinity (P. T. Jones et al. (1986), Nature 321:522). Overall, CDR-grafted antibodies consist of more than 80% human amino acid sequences (C. Queen et al. (1989), Proc. Natl. Acad. Sci. U.S.A. 86:10029; P. Carter et al. (1992), Proc. Natl. Acad. Sci. U.S.A. 89:4285). Despite these efforts, CDR-grafted, humanized antibodies were shown to still evoke an antibody response against the grafted V region (W. Y. Hwang et al. (2005), Methods 36:3).
Subsequently to CDR grafting, humanization methods based on different paradigms such as resurfacing (E. A. Padlan et al. (1991), Mol. Immunol. 28:489), superhumanization (P. Tan D. A. et al. (2002), J. Immunol. 169:1119), human string content optimization (G. A. Lazar et al. (2007), Mol. Immunol. 44:1986) and humaneering have been developed in an attempt to further decrease the content of non-human sequences in therapeutic mAbs (J. C. Almagro et al. (2008), Frontiers in Bioscience 13:1619). As in CDR grafting approaches, these methods rely on analyses of the antibody structure and sequence comparison of the non-human and human mAbs in order to evaluate the impact of the humanization process into immunogenicity of the final product. When comparing the immunogenicity of chimeric and humanized antibodies, humanization of variable regions appears to decrease immunogenicity further (W. Y. Hwang et al. (2005), Methods 36:3-10).
De-immunization is another approach developed to reduce the immunogenicity of chimeric or mouse antibodies. It involves the identification of linear T-cell epitopes in the antibody of interest, using bioinformatics, and their subsequent replacement by site-directed mutagenesis to human or non-immunogenic sequences (WO 9852976 A1, the contents of which are incorporated by this reference). Although de-immunized antibodies exhibited reduced immunogenicity in primates, compared with their chimeric counterparts, some loss of binding affinity was observed (M. Jain et al. (2007). Trends in Biotechnol. 25:307).
The development of phage display technology complemented and extended humanization approaches in attempts to obtain less immunogenic mAbs for therapy in humans. In phage display, large collections (“libraries”) of human antibody VH and VL regions are expressed on the surface of filamentous bacteriophage particles. From these libraries, rare phages are selected through binding interaction with antigen; soluble antibody fragments are expressed from infected bacteria and the affinity of binding of selected antibodies is improved by mutation (G. Winter et al. (1994), Annu. Rev. Immunol. 12:433). The process mimics immune selection, and antibodies with many different bindings specificities have been isolated using this approach (H. R. Hoogenboom et al. (2005), Nat. Biotechnol. 23:1105). Various sources of H and L chain V regions have been used to construct phage display libraries including those isolated from non-immune or immune donors. In addition, phage display libraries have been constructed of V regions that contain artificially randomized synthetic CDR regions in order to create additional diversity. Often, antibodies obtained from phage display libraries are subjected to in vitro affinity maturation to obtain high affinity antibodies (H. R. Hoogenboom et al. (2005), Nat. Biotechnol. 23:1105).
The creation of transgenic mouse strains producing human antibodies in the absence of mouse antibodies has provided another technology platform for the generation of specific and high affinity human mAbs for application in humans. In these transgenic animals, the endogenous mouse antibody machinery is inactivated and replaced by human Ig loci to substantially reproduce the human humoral immune system in mice (A. Jakobovits et al. (2007), Nat. Biotechnol. 25:1134; N. Lonberg (2005), Nat. Biotechnol. 23:1117). B cell development as well as Ig diversification by recombination of gene segments is faithfully reproduced in these mice, leading to a diverse repertoire of murine B cells expressing human Igs. By immunizing these mice with antigens, it was further demonstrated that these transgenic animals accumulated somatic mutations in the V regions of both heavy and light chains to produce a wide diversity of high-affinity human mAbs (N. Lonberg (2005), Nat. Biotechnol. 23:1117).
The question, whether “fully human” mAbs such as derived from phage display libraries or transgenic mice are less immunogenic than humanized mAbs cannot be answered yet, because full immunogenicity data are available for just two human mAbs. An anti-tumor necrosis factor mAb, developed from phage-displayed human libraries induced antibody responses in 12% of patients—at the higher end of the incidence of anti-antibody responses of the humanized antibodies (W. Y. Hwang et al. (2005), Methods 36:3-10).
Evaluation of the immunogenicity of the first registered human mAb generated by the transgenic approach demonstrated that mAb treatment resulted in the generation of antibodies in approximately 5.5% of treated cancer patients (A. Jakobovits et al. (2007), Nat. Biotechnol. 25:1134; J. A. Lofgren et al. (2007), J. Immunol. 178:7467).